Extract spatiotemporal effects from fitted model

Extract posterior distributions of spatiotemporal interaction effects from a fitted tulpa model.

Usage

spatiotemporal_effects(
  object,
  format = c("array", "long", "summary"),
  probs = c(0.025, 0.5, 0.975),
  ...
)

# S3 method for class 'tulpa_fit'
spatiotemporal_effects(
  object,
  format = c("array", "long", "summary"),
  probs = c(0.025, 0.5, 0.975),
  ...
)

Arguments

object

A tulpa_fit object fitted with spatiotemporal argument

format

Output format: "array" (default, S x T x draws), "long" (data frame with s, t, draw, value columns), or "summary" (posterior summaries).

probs

Quantiles to compute if format = "summary".

...

Ignored

Value

Spatiotemporal effects in requested format

Examples

if (FALSE) { # \dontrun{
# `fit` carries a Knorr-Held interaction block, which no tulpa backend
# fits: the spec comes from a model package that configures the
# interaction itself (see spatiotemporal() for what the engine does fit).
st_effects <- spatiotemporal_effects(fit, format = "summary")
head(st_effects)
} # }